favorite_count = sum of likes received for postsretweet_count = sum of retweets received for postsretweet_own_tweets_count = sum of retweets received for posts that were created by the botsfavorite_own_tweets_count = sum of likes received for posts that were created by the botsThe classification of accounts by language rely on the following logic:
If 66% of the tweets or more are in a given language, the account is classified to predominantly tweet in that language.
Example:
this account https://twitter.com/AngieKayman tweets are 96.95% classified by Twitter as German so I set account_lang to german.
Here I only classify accounts as English, Indonesian, Dutch and German because only for those languages are there enough tweets (the rest are probably misclassifations on behalf of Twitter anyway). If none of the accounts have above 66% tweets in one language, I code them as mixed. All others are classified as other (again probably missclassifacations but this number is small).
| account_lang | n |
|---|---|
| indonesian | 58 |
| english | 55 |
| dutch | 29 |
| mixed | 26 |
| german | 11 |
| other | 9 |
## [1] 53369
| lang | n | perc |
|---|---|---|
| en | 16179 | 30.32 |
| nl | 13265 | 24.86 |
| in | 12800 | 23.98 |
| de | 6407 | 12.01 |
| und | 4177 | 7.83 |
| tl | 190 | 0.36 |
| ro | 120 | 0.22 |
| ja | 81 | 0.15 |
| es | 20 | 0.04 |
| ca | 19 | 0.04 |
| da | 17 | 0.03 |
| ko | 17 | 0.03 |
| ht | 11 | 0.02 |
| it | 11 | 0.02 |
| fr | 10 | 0.02 |
| et | 8 | 0.01 |
| fi | 7 | 0.01 |
| pl | 6 | 0.01 |
| cy | 4 | 0.01 |
| hu | 3 | 0.01 |
| no | 3 | 0.01 |
| hi | 2 | 0.00 |
| ru | 2 | 0.00 |
| sv | 2 | 0.00 |
| tr | 2 | 0.00 |
| ar | 1 | 0.00 |
| eu | 1 | 0.00 |
| is | 1 | 0.00 |
| lt | 1 | 0.00 |
| lv | 1 | 0.00 |
| sl | 1 | 0.00 |
37k tweets (86% of all tweets) have no location. The rest do have a specified tweet location.
| location | n | perc |
|---|---|---|
| Papua | 1762 | 21.57 |
| Bekasi Barat, Indonesia | 719 | 8.80 |
| Raja Ampat | 697 | 8.53 |
| London, England | 673 | 8.24 |
| Teluk Bintuni | 507 | 6.21 |
| Hawaii | 326 | 3.99 |
| Birmingham | 319 | 3.91 |
| Brisbane | 315 | 3.86 |
| Australia | 303 | 3.71 |
| Amsterdam | 286 | 3.50 |
| Bojonegoro, Indonesia | 253 | 3.10 |
| London | 246 | 3.01 |
| Mojosari, Indonesia | 199 | 2.44 |
| Mataram, Indonesia | 171 | 2.09 |
| Semarang, Jawa Tengah | 171 | 2.09 |
| Papua Barat, Indonesia | 166 | 2.03 |
| Batu, Indonesia | 150 | 1.84 |
| Kupang, Kedah | 141 | 1.73 |
| Kota Makassar, Sulawesi Selata | 135 | 1.65 |
| Konawe Selatan, Sulawesi Tengg | 127 | 1.55 |
| Bali, Indonesia | 126 | 1.54 |
| Wamena | 97 | 1.19 |
| Biak | 94 | 1.15 |
| Amsterdam, The Netherlands | 68 | 0.83 |
| Kota Surabaya, Jawa Timur | 42 | 0.51 |
| Tanah Abang, Indonesia | 34 | 0.42 |
| Papua, Indonesia | 23 | 0.28 |
| Martapura, Indonesia | 19 | 0.23 |
| location | n | perc |
|---|---|---|
| Indonesia | 5633 | 68.96 |
| London/Amsterdam/Birmingham | 1592 | 19.49 |
| Australia | 618 | 7.57 |
| Hawaii | 326 | 3.99 |
## [1] 44462
| lang | n | perc |
|---|---|---|
| en | 14992 | 33.72 |
| nl | 12317 | 27.70 |
| in | 7286 | 16.39 |
| de | 6397 | 14.39 |
| und | 3050 | 6.86 |
| tl | 148 | 0.33 |
| ro | 99 | 0.22 |
| ja | 80 | 0.18 |
| da | 17 | 0.04 |
| ca | 16 | 0.04 |
| fr | 10 | 0.02 |
| es | 9 | 0.02 |
| et | 7 | 0.02 |
| fi | 7 | 0.02 |
| it | 6 | 0.01 |
| cy | 4 | 0.01 |
| pl | 4 | 0.01 |
| hu | 3 | 0.01 |
| sv | 2 | 0.00 |
| tr | 2 | 0.00 |
| eu | 1 | 0.00 |
| hi | 1 | 0.00 |
| is | 1 | 0.00 |
| lt | 1 | 0.00 |
| lv | 1 | 0.00 |
| sl | 1 | 0.00 |
| location | n | perc |
|---|---|---|
| Papua | 1762 | 27.44 |
| Raja Ampat | 697 | 10.86 |
| London, England | 669 | 10.42 |
| Teluk Bintuni | 507 | 7.90 |
| Brisbane | 283 | 4.41 |
| Hawaii | 266 | 4.14 |
| Australia | 258 | 4.02 |
| Birmingham | 254 | 3.96 |
| London | 245 | 3.82 |
| Amsterdam | 231 | 3.60 |
| Bekasi Barat, Indonesia | 137 | 2.13 |
| Mojosari, Indonesia | 108 | 1.68 |
| Batu, Indonesia | 104 | 1.62 |
| Kota Makassar, Sulawesi Selata | 103 | 1.60 |
| Mataram, Indonesia | 103 | 1.60 |
| Bali, Indonesia | 100 | 1.56 |
| Wamena | 97 | 1.51 |
| Konawe Selatan, Sulawesi Tengg | 89 | 1.39 |
| Kupang, Kedah | 88 | 1.37 |
| Amsterdam, The Netherlands | 68 | 1.06 |
| Biak | 68 | 1.06 |
| Bojonegoro, Indonesia | 50 | 0.78 |
| Papua Barat, Indonesia | 48 | 0.75 |
| Semarang, Jawa Tengah | 43 | 0.67 |
| Tanah Abang, Indonesia | 18 | 0.28 |
| Papua, Indonesia | 13 | 0.20 |
| Martapura, Indonesia | 8 | 0.12 |
| Kota Surabaya, Jawa Timur | 4 | 0.06 |
| location | n | perc |
|---|---|---|
| Indonesia | 4147 | 64.58 |
| London/Amsterdam/Birmingham | 1467 | 22.85 |
| Australia | 541 | 8.43 |
| Hawaii | 266 | 4.14 |
## [1] 8907
| lang | n | perc |
|---|---|---|
| in | 5514 | 61.91 |
| en | 1187 | 13.33 |
| und | 1127 | 12.65 |
| nl | 948 | 10.64 |
| tl | 42 | 0.47 |
| ro | 21 | 0.24 |
| ko | 17 | 0.19 |
| es | 11 | 0.12 |
| ht | 11 | 0.12 |
| de | 10 | 0.11 |
| it | 5 | 0.06 |
| ca | 3 | 0.03 |
| no | 3 | 0.03 |
| pl | 2 | 0.02 |
| ru | 2 | 0.02 |
| ar | 1 | 0.01 |
| et | 1 | 0.01 |
| hi | 1 | 0.01 |
| ja | 1 | 0.01 |
| location | n | perc |
|---|---|---|
| Bekasi Barat, Indonesia | 582 | 33.30 |
| Bojonegoro, Indonesia | 203 | 11.61 |
| Semarang, Jawa Tengah | 128 | 7.32 |
| Papua Barat, Indonesia | 118 | 6.75 |
| Mojosari, Indonesia | 91 | 5.21 |
| Mataram, Indonesia | 68 | 3.89 |
| Birmingham | 65 | 3.72 |
| Hawaii | 60 | 3.43 |
| Amsterdam | 55 | 3.15 |
| Kupang, Kedah | 53 | 3.03 |
| Batu, Indonesia | 46 | 2.63 |
| Australia | 45 | 2.57 |
| Konawe Selatan, Sulawesi Tengg | 38 | 2.17 |
| Kota Surabaya, Jawa Timur | 38 | 2.17 |
| Brisbane | 32 | 1.83 |
| Kota Makassar, Sulawesi Selata | 32 | 1.83 |
| Bali, Indonesia | 26 | 1.49 |
| Biak | 26 | 1.49 |
| Tanah Abang, Indonesia | 16 | 0.92 |
| Martapura, Indonesia | 11 | 0.63 |
| Papua, Indonesia | 10 | 0.57 |
| London, England | 4 | 0.23 |
| London | 1 | 0.06 |
| location | n | perc |
|---|---|---|
| Indonesia | 1486 | 85.01 |
| London/Amsterdam/Birmingham | 125 | 7.15 |
| Australia | 77 | 4.41 |
| Hawaii | 60 | 3.43 |
## [1] 188
## [1] 22181
## [1] 11450
Earliest tweet was posted in 2014
Let’s zoom in a bit. Looks like there is a clear spike in activity in mid June 2020. This overlaps with a lot of the accounts that were created on June 14th 2020.
Black bars show accounts that are not in the initial bot sample.
There are 282 identical tweets that were posted by more than 8 accounts. Some texts repeat more often but the hashtags section after the text seems to often change slightly.
Nodes (dots) are colored in by language of the account. Seems like interactions and follows are not really dependent on what language the accounts post in.